Sima.ai launches chips to run inference-based LLMS-on-device under 10 watts

AI News


Sima.ai, a US-based AI chip company, has launched a next-generation platform to accelerate the growth of physical AI applications. The company has introduced Modalix machine learning systems in its chip (MLSOC) along with new system-on-modules (SOM) and development kits, with the aim of meeting the growing demand for physical AI in sectors such as robotics, automotive, and healthcare.

Modalix is a second generation MLSOC designed to deliver high performance without compromising power efficiency that operates under 10 watts. “We're here at the age of physical AI. Now, while Modalix is in production, we're accelerating its global adoption,” said Krishna Rangasayee, founder and CEO of Sima.ai.

Flexible arm-based architecture enables seamless support for real-time decision-making, natural language interaction, large-scale language models (LLM), transformers, convolutional neural networks (CNNS), and generator AI (genai) workloads.

Seamless integration

Sima.ai has also announced its new Modalix SOM. This feature makes it easy to integrate into existing systems and makes it an attractive option for developers.

The platform includes MIPI, memory, and other essential I/O components for rapid scaling of physical AI systems. The company's LLIMA software framework, which supports LLM deployment, further simplifies integration of Genai features into physical AI applications.

Ami Badani, Chief Marketing Officer at ARM, said: “Sima.ai enables smarter and more sustainable systems across the industry.”

Sima.ai worked with Synopsys to speed up development using its AI-powered design tools. TSMC's advanced N6 process has allowed Modalix to meet power, heat and reliability standards.

Sajiv Dalal, president of TSMC North America, said, “This collaboration underscores our commitment to driving energy-efficient chip innovation, which is redefineing the future of AI.”

Global launch and commercial availability

New Modalix SOM and DevKit are now available. The 8GBSOM priced at $349, while the 32GB version costs $599. The DevKit price is $1,499. Sima.ai aims to accelerate the global adoption of physical AI with strong demand for products.

In a recent exclusive interview with targetRangasai also praised the Indian market, saying, “I believe India will become the next market manufacturer not only for global consumption but also for global consumption. It suggested that the funds will expand globally and be used to expand the physical AI platform.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *